Depression, anxiety, cognitive impairment and their association with clinical and demographic variables in people with type 2 diabetes: a 4-year prospective study

Objective To investigate depression, anxiety and cognitive impairment and their associations with clinical and socio-demographic variables in type 2 diabetes. Methods The Zung Self-Rating Depression-Anxiety Scale and Mini-Mental State Examination (MMSE) were administered at baseline and after 4 years to 498 consecutive patients, 249 non-insulin treated (NIT) and 249 insulin treated (IT), aged 40-80 years. Results At baseline, IT patients were older, had longer disease duration, higher HbA1c and did more glucose monitoring (p < 0.001, all) but their depression scores were lower than among NIT (p = 0.006), with no differences for anxiety or MMSE. After 4 years, 72 patients were lost to the follow-up, of whom 18 had died. 41 NIT had switched to insulin and increased BMI (p = 0.004), blood pressure (p < 0.001), retinopathy severity (p = 0.03) and microalbuminuria (p = 0.0045), but did not change their scores for depression, anxiety or MMSE. The remaining 171 NIT improved fasting glucose (p = 0.006), total cholesterol (p < 0.0001), triglyceride (p = 0.0026) and HbA1c (p = 0.0006). Despite increased prevalence of microalbuminuria and retinopathy (p < 0.0001, both), depression (p = 0.04) and MMSE (p = 0.0007) improved. Foot ulcers (p = 0.03), retinopathy (p < 0001), microalbuminuria (p = 0.0047) and hypertension (p < 0.0001) increased in the remaining 214 IT patients, in whom depression (p = 0.0005) and anxiety (p < 0.0001) worsened while MMSE improved slightly (p = 0.0002). On multivariate analysis, depression was associated with being a woman and anxiety with diabetes duration and lower schooling, which also affected MMSE scores. Conclusions Depression was associated with female gender and worsening complications but not modified by diabetes duration or switching to insulin therapy. Diabetes duration and lower schooling may affect anxiety and cognitive impairment.